摘要点击次数: 167
肖淑苹 (西安翻译学院 工程技术学院陕西 西安 710105) 
中文摘要:为了提高纹理图像识别的准确率,提出了一种经验模式分解和支持向量机(EMD-SVM)在纹理图像识别中的应用。首先采用经验模式分解对原始信号进行分解,得到一组固有模式函数;然后采用固有模式函数和残差之和构建特征子集,并用支持向量机对不同的特征进行纹理分类识别;最后对不同的自然纹理图像进行实验,并将结果与小波变换和支持向量机的纹理图像识别做了比较。实验结果表明, EMD-SVM的纹理图像识别率和识别精度高于小波变换和支持向量机的纹理图像识别。
中文关键词:经验模式分解  支持向量机  固有模式函数  纹理识别
Application of EMD-SVM in Texture Image Recognition
Abstract:The application of EMD-SVM in texture image recognition is proposed for texture recognition for improving texture image recognition accuracy rate.At first,It takes the empirical mode decomposition on the original signal and decomposes it into a set of Intrinsic Mode Function (IMF);Then take the IMF and the sum of residual to construct feature subset,using SVM to texture classification for different features.Finally,experiments are conducted on different natural texture images,also compared to wavelet transform and SVM.The experiment result show that the method proposed in this paper has the higher recognition accuracy rate.
keywords:EMD  SVM  IMF  texture recognition
查看全文   查看/发表评论   下载pdf阅读器